Re: [PATCH v6 1/2] mm: migration: fix migration of huge PMD shared pages

[Michal Hocko]

On Fri 24-08-18 11:08:24, Mike Kravetz wrote:
> On 08/24/2018 01:41 AM, Michal Hocko wrote:
> > On Thu 23-08-18 13:59:16, Mike Kravetz wrote:
> > 
> > Acked-by: Michal Hocko <mhocko@xxxxxxxx>
> > 
> > One nit below.
> > 
> > [...]
> >> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> >> index 3103099f64fd..a73c5728e961 100644
> >> --- a/mm/hugetlb.c
> >> +++ b/mm/hugetlb.c
> >> @@ -4548,6 +4548,9 @@ static unsigned long page_table_shareable(struct vm_area_struct *svma,
> >>  	return saddr;
> >>  }
> >>  
> >> +#define _range_in_vma(vma, start, end) \
> >> +	((vma)->vm_start <= (start) && (end) <= (vma)->vm_end)
> >> +
> > 
> > static inline please. Macros and potential side effects on given
> > arguments are just not worth the risk. I also think this is something
> > for more general use. We have that pattern at many places. So I would
> > stick that to linux/mm.h
> 
> Thanks Michal,
> 
> Here is an updated patch which does as you suggest above.
[...]
> @@ -1409,6 +1419,32 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
>  		subpage = page - page_to_pfn(page) + pte_pfn(*pvmw.pte);
>  		address = pvmw.address;
>  
> +		if (PageHuge(page)) {
> +			if (huge_pmd_unshare(mm, &address, pvmw.pte)) {
> +				/*
> +				 * huge_pmd_unshare unmapped an entire PMD
> +				 * page.  There is no way of knowing exactly
> +				 * which PMDs may be cached for this mm, so
> +				 * we must flush them all.  start/end were
> +				 * already adjusted above to cover this range.
> +				 */
> +				flush_cache_range(vma, start, end);
> +				flush_tlb_range(vma, start, end);
> +				mmu_notifier_invalidate_range(mm, start, end);
> +
> +				/*
> +				 * The ref count of the PMD page was dropped
> +				 * which is part of the way map counting
> +				 * is done for shared PMDs.  Return 'true'
> +				 * here.  When there is no other sharing,
> +				 * huge_pmd_unshare returns false and we will
> +				 * unmap the actual page and drop map count
> +				 * to zero.
> +				 */
> +				page_vma_mapped_walk_done(&pvmw);
> +				break;
> +			}

This still calls into notifier while holding the ptl lock. Either I am
missing something or the invalidation is broken in this loop (not also
for other invalidations).

-- 
Michal Hocko
SUSE Labs